Procedure
Main steps
An ALM study is the ideal instrument for the Board of Trustees to make investment decisions on an informed basis. The study is based on a dynamic projection of the whole portfolio of insured per-sons in the future.
The core task of asset liability management is the risk management of invested fund assets with regard to fulfilling the promised benefits in terms of costs and time. At the centre of risk manage-ment is the identification, evaluation and targeted assumption of investment risks.
The diagram below shows the impact factors from the asset and liability perspective. The ALM study takes all the variables into account. In addition to a risk evaluation, an investment proposal is also always put forward for each portfolio development scenario.
Main steps in preparing the study:
- Investigation of specific impact factors
- Projection of the portfolio of insured persons
- Calculation of the financial and structural risk capacity
- Calculation of the dynamic targeted returns for a stable cover ratio development
- Calculation of the target returns taking into consideration the formation of the value fluctuation reserve and the benefit targets
- The compatible investment strategy (robuste Markowitz-Portfolio-Optimierung, welche die Un-terschreitungswahrscheinlichkeit sowohl der Zielrendite als auch der Sanierungskapazität mini-miert)
- Test of the possible portfolio in historical and synthetic extreme situations under the Swiss Sol-vency Test (stress tests)
The investment strategy is determined as a result of a structured and documented process.
Calculating the risk capacity
Overall risk capacity is assessed from the following perspectives:
We take an in-depth look at structural risk capacity and its development over time. Key data is collected on, for example, vested pension capital, ratio of active insured persons to pensioners, reserves, dynamic targeted return and restructuring capacity. These calculations are made for every projection year.
We also look at financial risk capacity aspects, determined by the cash flow situation as well as the available value fluctuation reserve.
We are thus able to show the changes in overall risk capacity over time in the form of a graph.
Calculation of expected returns
The value of the assets develops stochastically on the basis of a Monte Carlo simulation.
One of the most important factors for evaluating different asset allocations is the estimation of ex-pected investment return and its volatility.
The starting point for the projected returns in all asset classes in any currency zone is the antici-pated development of the respective yield curve according to the particular rating class.
Unsere Methode zur Herleitung der erwarteten Renditen unterscheidet zwischen zinssensitiven und nicht zinssensitiven Anlageklassen. Währenddem sich Marktwertveränderungen und Erträge von zinssensitiven Anlageklassen vollständig durch Veränderungen der Zinsstrukturkurven erklären lassen, sind nicht (vollständig) zinssensitive Anlageklassen oft volatiler und weisen gegenüber risi-koarmen Anlageklassen eine entsprechende Risikoentschädigung aus.
Interest-sensitive asset classes
The current yield curve contains market information on future interest rate expectations. We calcu-late the value of each individual bond in the bond portfolio or benchmark looking forward year to year based on the expected yield curve. The anticipated changes in market value and the coupons enable the anticipated returns to be calculated.
The use of yield curves allows differentiated consideration of how the various bond portfolios be-have under different market conditions. If currency hedging is carried out for example, this detailed approach enables hedging costs to be estimated systematically.
The market value of interest-rate-sensitive instruments (such as bonds and mortgages) reacts strongly to interest rate changes. An interest rate rise leads to a fall in the market value and vice versa. Thus an interest rate rise initially reduces return, before a positive impact is gradually felt as a result of the issue of bonds with higher coupons.
These zeitlich verzögerte changes in value von zinssensitiven Anlageklassen cannot be recorded using the commonly applied risk premium approach. With the risk premiums approach for interest-rate-sensitive asset classes, an interest rate rise wrongly leads to higher expected returns although the market value of the assets already in the portfolio initially falls sharply
Non-interest-sensitive asset classes
Asset classes, whose price development is not exclusively based on interest rate development, are assessed using the risk premium concept (e.g. equity).
In line with an economic downturn, the interest rate for short-term assets is often reduced, so that lower equity returns are also expected in a difficult free market environment with low interest rates. This effect is much less pronounced for long-term interest rates (e.g. 10-year federal bonds). For this reason it makes sense to gear the addition of risk premiums to short-term interest rates.
Our method also takes into consideration the different interest rates in the various countries. As a result of the no-arbitrage principle, equity returns from foreign companies should be identical after hedging of the currency risk. The investor accordingly requires additional risk compensation for bearing the currency risk.
Comprehensive risk management
We regard ALM as a comprehensive form of risk management. In addition to defining an opti-mised investment strategy, all other variable factors and plan parameters (e.g. technical interest rate or conversion rate) can be checked for their effectiveness. This allows us to identify the drivers for best achieving the employee benefits institution’s objectives.
Our ALM model allows us to carry out impact analyses of restructuring concepts. It gives us an-swers regarding the optimal combination of restructuring measures in order to minimise costs for employers and beneficiaries. Our ALM functionalities thus facilitate the implementation of efficient risk management from the perspective of employee benefits institutions and the company.
Depending on the individual needs of an employee benefits institution, we also analyse other spe-cific issues which may go beyond the scope of a conventional ALM study.
Stresstests
The 16 stress tests which are carried out on the portfolio and the proposed optimised portfolios are based on the provisions of the Swiss Solvency Test (SST). Simulations are made regarding the behaviour of the current and optimised portfolios during previous crises and under theoretical sce-narios such as inflation, deflation, equity and real estate crises, etc.
Portfolios exhibit fundamentally different behaviour depending on the scenario and the breakdown of currencies, corporate vs. government bonds, rating classes and maturities.
This gives the Board of Trustees an insight into the behaviour of the portfolio under a wide range of market-related extreme conditions.
The graph below shows the impact of stress tests on different portfolio samples.
Optimisation of the investment strategy
Optimisation of the investment strategy is based on the Markowitz method. Key impact factors are expected return, volatilities and correlations. Since these parameters are traditionally very sensitive to optimisation and are difficult to estimate, a sophisticated accounting procedure is applied. All potential portfolios are of course checked for conformity to the BVV 2. The fund can also set addi-tional limitations to the portfolios (e.g. universe, minimum and maximum limits, relative ratios be-tween certain asset classes and maximum loss in one or more stress scenarios). The decisive fac-tor for the pension fund’s selection of the optimal portfolio from the set of efficient portfolios is ulti-mately the target or targeted return and the risk or restructuring capacity of the fund. Darüber hin-aus kann der Stiftungsrat eine Hierarchie von weiteren Zielen definieren.
Example of the efficient frontier and the selection of the optimal asset allocation accounting for the utility optimization:
